One of the most powerful tools we offer is our BatchRank propensity score (likelihood-to-sell). It helps you zero in on your most promising leads, so you can focus your time and resources where they'll have the biggest impact.
The Basics
When a 4.0 user selects “off market” as their strategy, they will be served a button called “BatchRankAI” When the user clicks this button, they can use the slider to surface properties that have a low, medium, or high BatchRank score. The higher the score, the greater the likelihood that a property will sell in the near future.
In My Lists, there is a BatchRankAI column. This column shows whether a property has a low, medium or high BatchRank score. Again, the higher the score, the greater the likelihood that a property will sell in the near future.
Explaining the Technology
BatchRank is built using sophisticated machine learning algorithms. We use both supervised and unsupervised machine learning models.
Supervised Learning: Think of it like having a teacher. The computer is given examples with the right answers, and it learns to make similar predictions on new data.
Unsupervised Learning: Think of it like exploring on your own. The computer finds patterns and groups in data without being told what to look for.
Supervised learning helps us predict which leads will convert based on past successes. Unsupervised learning helps us uncover hidden patterns in your data to gain even deeper insights.
We analyze a vast array of data points (Over 800 data points) from our big data models that we make available to our enterprise clients (www.batchata.com). This includes data such as assessor and recorder data (who owns properties and how they are financed), demographic information (net worth, age), building permits, real time MLS data, AVM models, market trends, and other distress indicators such as involuntary liens, pre-foreclosure, vacant flags (USPS data), among others.
Think of BatchRank as having a crystal ball that predicts which properties are most likely to sell (both on and off-market).
Highlighting the Benefits
Targeted Opportunities - By prioritizing leads with high BatchRank scores, you can dramatically improve your close rates
You'll waste less time chasing leads that are unlikely to convert, and you'll spend more time nurturing relationships likely to lead to a contract
This translates to shorter sales cycles, higher close rates, and ultimately, increased revenue for your business.
Competitive Advantage: Our model helps you stay ahead of the curve. You'll have access to insights that other agents might miss, giving you a real edge in the market.
Addressing Concerns/Questions:
Is it accurate?
We've invested heavily in developing and refining the model. It's constantly learning and improving, and we're seeing excellent results in terms of predicting property sales.
How does it work exactly?
It's a complex process involving advanced algorithms, but the key is that it learns from historical data to identify patterns and make predictions.
Can users layer additional filters on top of BatchRank?
Yes! Stack filters like pre-foreclosure, years owned etc with BatchRank to identify properties that best fit your niche.
What data is the model trained on
The model is currently trained on our nationwide property data consisting of roughly 155 million property parcels. We have an even more advanced model coming soon that will be trained at the locality level (think county/metro areas). This model will adapt to real-time market conditions in the local market and achieve a higher accuracy with likelihood to sell predictions. Real estate markets are localized and this model will outperform anything currently available in the marketplace.